I run a B2B SaaS startup @harmonizehq and tweet about startup life and practical wisdom in books.
Delhi
Joined December 2013
- Tweets6K
- Following533
- Followers2.2K
- Likes4.9K
Most Indian family run companies are basically optimised around the promoter..
i've never worked for one but ppl close to me do & its always the same story: smartest guy in the room is always the CEO, lots of agreeable mid managers, ceremonial processes & filtered bad news..
#WATCH | How Did eMudhra’s Key Generation Flaw Slip Past Years Of Controls And Audits?
Kaushik Srinivasan of eMudhra, explains how a flaw in the April 2025 key-generation process led to weak cryptographic keys being exposed, triggering action by major browsers. He addresses what went wrong, the gaps in the process and the potential business impact.
@eMudhra @SnehiShah11
#eMudhra #CyberSecurity #DataSecurity #ETNOW
Have the utmost respect for people who pay a pretty penny to go to watch sports events in India, be inconvenienced and humiliated, just to support their teams.. Really big hearted people 🙏
Shame on @IndianFootball .
For 1000/- look at the condition. They didn't even clean the place
@7negiashish
@MarcusMergulhao
#IndianFootball
TIL - Tuberculosis has killed ~1/7th of all people who ever lived 🤯
I used to think of it as smallpox, plague, cholera.. stuff belonging to a different age but it killed 1million+ people in 2024!
Everything is Tuberculosis by John Green is a fascinating read..
an exhibit of why the powers that be don't really care about us
most people have made peace with "2-3x/month problems" & 've become sub humans
when somebody complains, they pull them down cause it hurts them to see a full human with self worth not willing to accept the problem
Amazing thread on our rich historical heritage & how little we've cared for them or encased them in opulence
as @1Force says "when our Gurudwaras were kuchcha (weak) our faith was pukka (strong) but now that are Gurudwaras are strong, the faith is weakens"
(i'm paraphrasing)
Watching Odyssey was made even more fun as bits from @DalrympleWill and @tweeter_anita pod on the collapse of the bronze age made their tiny appearances..
I chuckled a bit when they mentioned sea people, trade routes, even the "our bronze age is collapsing" in the end ..
Qwen models are terrific to run on local machines (provided you can spare 20 gigs)
I use qwen3.6-35b-a3b to have it rate my photos & give me feedback on framing, exposure and what not. All without my pictures (featuring kid, dog, wife et al) ever going to the internet.
Ujjwal retweeted
Qwen3.8-27B is coming! 🔥
Will run locally on 17GB RAM/VRAM setups.
📢Meet Qwen3.8-Max — our most capable model to date.
Next week, the open weights of Qwen3.8-Max will be released, and Qwen3.8-27B is also going open-weights to meet you all!🎉
Qwen3.8-Max, a new bar for coding and cowork at 2.4T parameters:
- Autonomous coding: 10+ days of self-evolving development, from empty folder to production without hand-holding, complete project trace in the GitHub：github.com/qwen-code-dev-bot…
- Real work, real results: Production-quality deliverables across hundreds of professions.
- Long-horizon mastery: System-level autonomous planning with closed-loop adaptive learning, driving 500+ turns of chip design optimization and 365 days of e-commerce strategy.
- Native multimodal intelligence: Vision isn't just input — it's a continuous feedback loop for planning, execution, and self-correction.
💰Pricing:
Input: $2.0 / M tokens
Output: $6.0 / M tokens
Implicit Caching: $0.25 / M tokens
Start building with Qwen3.8-Max! 🚀
📖 Blog: qwen.ai/blog?id=qwen3.8
✅ Qwen Studio: chat.qwen.ai/?models=qwen3.8…
⚡ API: qwencloud.com/models/qwen3.8…
I will try this.. the only issue is I hate my voice
One pattern I find useful for working with LLMs is a nice long ramble session. Sometimes the LLM needs more bits to understand what you're trying to achieve, but you're too lazy to type them. In these cases I like to lean back, switch to /voice and just ramble for like 10 minutes, total mess, anything goes, full stream of consciousness. Sometimes I declare it up top, something like "switching to speech recognition sorry for any typos...". Sometimes I turn it into a small interview of a few turns. But I find that the LLMs are somehow very good at reconstructing long incoherent rambles and often their echo of your own tangle of thoughts comes out quite a bit cleaner than what you started with. The result is that you improve the mind meld and have to correct things less from that point on.
insane transition happening in my trade.. there's so much variation in costs across models that software engineering is increasingly becoming an economics problem..
good engineers won't just do it faster and better, they'll do it cheaper too
times we live in.. line between legit product launches and april fool's day gag products are blurring..
Supremely underrated.
In Indian context, reading about most Padma Bhushan winners just blows ones mind. If we bump into most of these Padma Bhushan winners, they appear ordinary by any measurable contemporary metric but have changed a country of 1B people for the better..
Seneca dropped a mic 2000 years ago and it hasn't been picked since
"You map out what is in fortune's hand but let
slip what's in your own hand."
Highly recommend reading "On the shortness of life" dn790008.ca.archive.org/0/it…
If token prices do fall, its a compelling opportunity to use them to distil/fine tune much smaller custom models that run locally
Qwen3.6 35b on my mac kicks ass
is it anywhere as good as ant/openai, no?
does it do 80% of tasks, well enough, heck yeah!
moneycontrol.com/artificial-…
Recently read Angel Down that won the Pulitzer this year for Fiction.. I picked it up cause it was a ww1 story but it just messed with my head in different creative ways..
May be that's what art is supposed to.. take you weird places. Recommended, if you want something different